OplogToRedis: Run writer routines in parallel for each redis client. #75
+83
−42
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
lib/oplog/tail.go
Outdated
| @@ -116,7 +116,7 @@ func init() { | |||
|
|
|||
| // Tail begins tailing the oplog. It doesn't return unless it receives a message | |||
| // on the stop channel, in which case it wraps up its work and then returns. | |||
| func (tailer *Tailer) Tail(out []chan<- *redispub.Publication, stop <-chan bool) { | |||
| func (tailer *Tailer) Tail(out [][]chan<- *redispub.Publication, stop <-chan bool) { | |||
There was a problem hiding this comment.
this 2d array of channels is getting complicated-enough that the semantics aren't obvious. could we document it, or use type aliases, or something similar that would make it easier to understand?
main.go
Outdated
|
|
||
| waitGroup.Add(1) | ||
|
|
||
| // We crate two goroutines: |
There was a problem hiding this comment.
lol that was there, i'll do a spelling pass
| redisPubsAggregationEntry := make([]chan<- *redispub.Publication, clientsSize) | ||
| stopRedisPubsEntry := make([]chan bool, clientsSize) | ||
|
|
||
| for j := 0; j < clientsSize; j++ { |
There was a problem hiding this comment.
general comment, this has grown somewhat hard to follow because of the nested arrays and loops, etc. it would be nice to refactor it to be a bit more abstract when we're back on firmer ground.
There was a problem hiding this comment.
I think type aliases will help with this somewhat, but will also add some comments.
torywheelwright
approved these changes
May 2, 2024
|
(small reminder to bump version before merging) |
alex-goodisman
force-pushed
the
alex.sentinel-parallel
branch
from
12f7215
to
c7d89ce
Compare
alex-goodisman
added a commit
that referenced
this pull request
May 3, 2024
Revert #75 Also bump to 3.5 for all the parallelism & orchestration changes
alex-goodisman
added a commit
that referenced
this pull request
May 13, 2024
Redo #75 in main, count through each redis client and make a separate chan and coroutine for clientsCount * ordinalCount different entries. make some changes to aggregation in order to accomodate this. in the tailer, get the [][]chan and for each incoming publication, first index by the shard ordinal (hash of db name), then iterate through the remaining chans (should be one for each redis client) and publish to all of them. at present, the publisher stream still accepts an array, but now it's only getting one (each). [fixed](66e79df) array index out of bounds issue
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
in main, count through each redis client and make a separate chan and coroutine for clientsCount * ordinalCount different entries. make some changes to aggregation in order to accomodate this.
in the tailer, get the [][]chan and for each incoming publication, first index by the shard ordinal (hash of db name), then iterate through the remaining chans (should be one for each redis client) and publish to all of them.
at present, the publisher stream still accepts an array, but now it's only getting one (each).